SemimetricPCA: SemimetricPCA

In sipemu/Nonparametric-Functional-Data-Analysis: Nonparametric Functional Data Analysis

Description

SemimetricPCA is a function that calculates a distance matrix based on the PCA of the functional data. A Rcpp port of the PCA-type semimetric function of Ferraty and Vieu.

Usage

SemimetricPCA(Data1, 
              Data2, 
              q, 
              EigenVec = NULL)

Arguments

Data1	Matrix with functional data (curves) each row one; n rows
Data2	Matrix with a second set of functional data (curves) each row one; m rows
q	Number of principal components
EigenVec	The Eigenvectors based on DATA1

Value

Semimetric returns a list with: semimetric as the n x m - distance-matrix EigenVec The Eigenvectors based on DATA1, which can be reused in following semimetric calculations

Author(s)

Simon Mueller [e-mail]

References

http://www.math.univ-toulouse.fr/staph/npfda/

Ferraty, F. and Vieu, P. Nonparametric Functional Data Analysis. Springer 2006.

See Also

Semimetric

Examples

# functional data sets
require (fds)

# fat spectrum dataset
X <- t(Fatspectrum$y)

# setup semimetric parameter
q <- 4

# calculate distance matrix
Dist1 <- SemimetricPCA (X[1:100, ], 
                        X[101:150, ], 
                        q)

# calculate distance matrix using Hhalf
Dist2 <- SemimetricPCA (X[1:100, ], 
                        X[151:215, ], 
                        q, 
                        Dist1$EigenVec)

sipemu/Nonparametric-Functional-Data-Analysis documentation built on May 29, 2019, 10:10 p.m.